New spelling may rool OK?

Last month, the English Spelling Society provisionally endorsed a new spelling system which it hopes will ultimately eventually replace the highly irregular system used today in spelling English. The new system is called Traditional Spelling Revised (TSR for short).

The Society believes that adopting TSR would help children and students to predict pronunciation from spelling, and so help them learn to read more quickly and with less effort, and to increase literacy levels.

The Society plans to review in 5 years whether the new spelling system is becoming an acceptable alternative to traditional spelling within the English Speaking World, and whether TSR needs change.

The Society’s materials suggest that TSR would respell around 25% of common words and around 10%-20% of the words in texts.

Disappointingly, the Society’s website does not yet highlight its provisional endorsement of TSR. News of the endorsement is tucked away in a news release and other information about TSR is hard to find. Also, I could find no full description of TSR. Here are 3 links to material I found on the Society’s website.

The English Spelling Society
Set up in 1908, The English Spelling Society was set up in 1908 to find solutions to bring about reform of the highly irregular English spelling system. The Society believes this system causes significant economic and social costs that would not exist if English had a more phonetic spelling system. Founding members of the Society included Andrew Carnegie, the Pitmans and George Bernard Shaw. Shaw (who wrote in Pitman’s shorthand) left money in his will to advance spelling reform.

What TSR would change

The main changes are that TSR would:

  • retain 2 important and pervasive spelling rules, but apply them more consistently. They are the doubling rule and the ‘magic e’ rule.
  • for vowels, each grapheme would stand for only one sound (one phoneme). But, in many cases, a phoneme could still be spelled using different graphemes in different words.
  • introduce 2 new  letter combinations: <aa> and <uu>.
  • remove some redundant letters. (In a few cases, an apostrophe would show that a redundant letter has been removed.)  
  • introduce a hyphen or apostrophe to help readers predict the pronunciation of some vowel combinations.

Graphemes and phonemes
A grapheme is a single letter, or sequence of 2 letters (digraph) or, occasionally, 3 letters (trigraph) used to represent a sound. A phoneme is a unit of sound that can distinguish one word from another word in the same language.

Convention places graphemes between <angled brackets> and phonemes between /slanted lines/.

Doubling rule

If a stressed vowel precedes a consonant and another vowel follows the consonant, traditional spelling generally writes the consonant as:

  • a single letter if the vowel preceding the consonant is pronounced long (eg <biter>, <sadist>).
  • a double letter if that vowel is short (eg <bitter>, <saddest>).

TSR would keeps the doubling rule, but would apply it more consistently. Table 1 shows some common words in which TSR would respell single consonants as double consonants.

governmentguvvernment*record [as noun]reccord
individualindividdualuse [as noun]uess [as noun]*
Table 1. Examples of consonants doubled in TSR

An Asterisk in Table 1 indicates that the spelling of the word also changes for another reason.

Table 2 shows some common words in which TSR would respell double consonants as single consonants.

Table 2. Examples of consonants no longer doubled in TSR

Magic e

Traditional spelling has a spelling rule that schools often teach to children as ‘magic e’. ‘Magic e’ is the grapheme <e>, used in single-syllable words after a consonant. The <e> is not pronounced, but tells the reader that the vowel preceding the consonant is long. Examples of words with and without ‘magic e’ are: <pane> / <pan> and <rote> / <rot>.

TSR would keep this spelling rule, but apply it more consistently. Thus, TSR would delete silent <e> where it is not the ‘magic e’, for example:   

  • <bloo> (instead of <blue>)
  • <elss> (instead of <else>)
  • <figgur> (instead of <figure>)
  • <inclood> (instead of <include>)
  • <involv> (instead of <involve>)
  • <liv> (instead of <live> when used as a verb, but for the noun still <life> [singular] and <lives> [plural])
  • <luv> (instead of <love>)
  • <mor> (instead of <more>)
  • <proov> (instead of <prove>)
  • <stor> (instead of <store>)

Both the ‘magic e’ rule and the doubling rule treat the cluster consonant followed by <l> as a single consonant.

New graphemes

TSR would create 2 new graphemes, both composed of double letters and referring to vowels for which traditional spelling provides no grapheme:

  • <uu> for the short vowel in words such as <good>. So, <foot>, <good >, <look> and <stood> would become <fuut>, <guud> <luuk> and <stuud>.
  • <aa> for long <a>. So, <father> would become <faather>.

One reason why English spelling does not match well with current pronunciation is the Great Vowel Shift


TSR would make several other changes to vowels. Those changes would eliminate alternative pronunciations for some graphemes. Thus, each grapheme would have only one pronunciation. But, for many phonemes, there would still be more than one way of spelling it. Said differently:

  • each grapheme would correspond to only one phoneme (sound).
  • more than one grapheme could still correspond to a single phoneme.

For example:

  • the grapheme <ow> would only spell the vowel written as <ow> in <town>. It would no longer spell the different vowel in <low>.
  • the vowel in <town> could still be written as <ow> (as in in <town>) or as <ou> (as in <proud>).

Here is a summary of the main changes to the vowels:

Long vowels

Firstly, to show long vowels:

  • <oo> would replace <u> for the long vowel in <blue>, <include> and <truth> (to become <bloo>, <inclood> and <trooth>)
  • <oo> would also replace <o> for the same long vowel in <prove> and <movie> (becoming <proov> and <moovee>)
  • <ee> would replace <e> for the long vowel in <period> (becoming <peeriod>) and for the long 2nd vowel in <experience> (to become <expeeri-ence>).
  • <ee> would also replace <ie> for the same long vowel in <believe> and <movie> (becoming <believe> and <moovee>)

Changes to use of <o>

Secondly, <o> would no longer spell the vowels in:

  • <company>, <cover>,  <government>, <love>, <money>, <month> and <worry> (becoming <company>, <cuvver>,<guvvernment>, <luv>, <munny>, <munth> and <wurry>)
  • <only> (which would become <oanly>) and <control> (becoming <controal>)

Thirdly, <o> would replace:

  • <ou> for the vowel in <shoulder> and <source> (becoming <sholder> and <sorce>). But <soul> would become <soal>; the reason for this isn’t stated, but I presume it is to avoid confusion with <sole>.
  • <ow> for the vowel in < low> and <show> (to become <lo> and <sho>). But in plurals (eg <flows> and past tenses (eg <slowed>, <oe> would be used (producing <floes> and <sloed>.
  • <a> for the vowel in <wash> and <quality> (to become <wosh> and <quollity>)

Changes to use of <e>

Fourthly, <e> would replace:

  • <ea> for the vowel in <head> and <health> (becoming <hed> and <helth>).
  • <ea> as part of the grapheme <ear> for the vowel in <learn> (to become <lern>)

Other changes to graphemes for vowels

Fifthly, <ai> would replace:

  • <ea> for the vowel in <bear> (to become <bair>)
  • <a> for the vowel in <bass> [as in bass guitar] (to become <baiss>)


There would be some changes to the graphemes for consonants, though not nearly as many as for vowels.

I don’t try to summarise here what the changes are for consonants because the TSR materials provided on the ESS website contain little discussion of this, other than deleting some redundant letters.

Redundant letters

TSR would remove some letters from words where they serve no purpose. Examples include <bild> (instead of <build>), <campain> (instead of <campaign>), <chainge (instead of <change>), <nash> (instead of <gnash>), <nolledge> (instead of <knowledge>), <our> (instead of <hour>) and <rong> (instead of <wrong>).

Also, <gh> would be dropped from some words (<altho> and <throo> instead of <although> and <through>) or replaced by <f> (<enuf> instead of <enough>) or <ff> (<coff> instead of <cough>)

If removing a redundant letter could make the meaning unclear, TSR would use devices such as:

  • an apostrophe to indicate that a traditional letter has been omitted [eg <‘our> (unit of time) versus <our> (possessive adjective); <‘no> (know) versus <no> (negative); or <h’art> (cardiac organ) versus <hart> (type of deer)] 
  • double letters [eg <sun> (celestial object) versus <sunn> (male child)]
  • additional letters [eg <bloo> (colour) versus <blooh> (past tense of <blow>)]


A hyphen would be inserted in between the constituent letters of some vowels to show that they are sequences of 2 graphemes and not a digraph (a single grapheme pronounced as a single vowel sound). For example, <create>, <experience>, <realise> and <theory> would become <cre-ate>, <expeeri-ence>, <re-alise>; and <the-ory>.

What TSR would not change

TSR would:

  • not introduce any new letters or any diacritics that modify existing letters (though it would introduce the 2 new digraphs <aa> and <uu>)
  • not change the spelling of a few particularly common words whose current spelling conflicts with the main principles underlying TSR
  • not change the spelling of some suffixes and some other familiar letter combinations whose pronunciation is usually predictable, even though their pronunciation conflicts with TSR’s main principles
  • not change the spelling of most unstressed vowels
  • not require the respelling of proper names (people, places, organisations), though respelling would be permitted. So, TSR would permit respelling <John>, <London> and <Belize> as <Jon>, <London> and <Beleez>, but would not require those changes.
  • not change the spelling of foreign words borrowed into English. TSR would italicise those words to show that the TSR spelling conventions do not apply to them.

Common irregular words

The following very common words would retain their original spelling:

  • personal pronouns and adjectives, and interrogative pronouns: <I, you, he, she, we, me, us, they, your, their(s), them, what, where, who, whose>
  • parts of the verbs to be and to have: <are, was, were, have, having>
  • numbers: <(n)one, once, two, four, seven, eleven, twelve, fourteen, seventeen>
  • days, months and seasons of the week: <Monday, Wednesday, Saturday, January, February, April, July, autumn>
  • the definite article, and demonstrative adjectives and adverbs: < the, than, that, then, thence, there, these, this, those, thus>
  • miscellaneous: <of, to(day), any(one), (n)either, nothing, some, woman, women, yes>.

Retaining some existing combinations

TSR would retain existing spellings of:

  • some suffixes, including <-tion>, <-ssion>, <-cial>, <-cious>, <-cean>, <-sion> and <-sure>
  • some letter combinations, such as <-alk>, <-aste>, <-ign>, <-ind>, <-old>, <-olk>, <-other>, <war->, <wor->
  • the combination <-ould> for the vowel in the modal verbs <could>, <should> and <would>. But in all other words, the new grapheme <uu> would spell that vowel.

TSR in different varieties of English

TSR seeks to provide common graphemes for all the main phonemes of spoken English in a manner that will permit speakers of different varieties of English to use their own traditions in pronouncing the sounds that a grapheme stands for.

  • In some cases, American spellings differ from British spellings. In many such cases, TSR would use the American spelling. For example, <-or> over < -our> (labor), <-ize> over < -ise> (theorize) and <-yze> over < -yse>, (analyze), <-er> over <-re> (center), <-og> over <-ogue> (catalogue).
  • As is already the case today, TSR would not use different graphemes to reflect regional differences in matters such as how to pronounce unstressed vowels, how to pronounce the vowels in <lot> or in <bath> and whether to pronounce /r/.
  • In words such as <duty>, <sue> and <tune>, some speakers (including many in North America) pronounce the stressed vowel as /u:/. Other speakers (including many in Britain) pronounce that vowel as /ju:/. To alert speakers to the varying pronunciation, TSR would insert an apostrophe after <u> in these instances. Thus <du’ty> (duty), <su’> (sue), and <tu’n> (tune).
  • TSR would permit divergent spellings for a small number of words where pronunciations differ too much to permit a common grapheme: <tomato> / <tomaato>; <vase>/ <vaaz>  vase); <booy> / <b’oy>.


Would adopting TSR a good idea? That question leads to 2 sub-questions:

  • is TSR better than the spelling system we use today?
  • is TSR  likely to be implemented?

Is TSR better than the spelling system we use today?

Many of the changes proposed in TSR have some merits. If TSR were to be implemented throughout the English-speaking World, English spelling would be much more consistent and much easier to learn.

Nevertheless, I think a few of the changes would not be a good idea:

  • the new grapheme <uu> for the vowel in <good>. I understand why TSR picked this grapheme: it avoids introducing a diacritic mark to modify the letter; and no other available existing or new digraph was obviously suitable. But all other digraphs composed of repeated vowels stand for long vowels, whereas the vowel in <good> is short. I don’t have a better solution to offer, but undermining a broad and logical spelling convention in this way seems unhelpful.
  • the new grapheme <aa> for the 1st vowel in <father>. Although this change would be logical, it would—according to the TSR materials—affect only a small number of words. For only a few words, it doesn’t seem worth introducing a new grapheme—even one that is logical.
  • replacing redundant letters with an apostrophe, eg replacing <know> with <’no> and <heart> with <he’rt>. Removing the redundant letter would: (a) enable readers to predict the pronunciation of such words; and (b) continue to distinguish these words from their homonyms—words pronounced in the same way, such as <no> and <hart>.
    But writers would need to remember that the spelling of these words includes an apostrophe and readers would still need to be able to recognise the old spellings for many decades while old books and other texts still circulate. So, overall, if there’s a need to continue to distinguish these words from their homonyms, I’m not convinced that changing the spelling is worthwhile.    
  • introducing an apostrophe in words like <tune>, <duty> and <news>to signal the fact that some varieties of English pronounce this vowel as /u:/ and other varieties pronounce it as /ju:/.       

Is TSR likely to be implemented?

Spelling reforms have been carried out for many languages, but for such reforms to succeed, they need to be imposed by a government or carried out by some other body whose lead people will follow. Even when backed by governments or by other persuasive organisations, spelling reforms have often generated controversy.   

For a spelling reform to gain much traction across the English-speaking world, it will need to be adopted widely in at least the US or the UK—and probably both. It seems inconceivable that any current or future US or UK governmental body would mandate a new spelling system for all purposes. English-speaking countries also have no private body (like the Académie Française) with enough authority to promulgate a new spelling system.  

If no government (or some other official authority) will require—or, at least, encourage—spelling reform, reform could happen only if a critical mass of individuals and organisations start switching to the news spellings. But no individual or organisation will do this by themself:

  • Individuals would suffer disadvantages (losing marks in exams, being regarded as illiterate, missing out on job opportunities) and would gain no benefits.
  • Businesses and educational institutions would suffer costs and perhaps other disadvantages but gain no benefits.
  • Publishers of dictionaries would incur costs. They would need to add the new spellings, and probably also have to keep the old spellings. And it is doubtful whether they would gain enough new customers to outweigh those costs.

In addition, old spellings would remain in existence in old books and other documents. At least some part of the population would need to retain at least some passive ability to read them, probably for many years. Realistically, it is not feasible to reprint all (or, probably, even many) old books and documents with updated spelling.

Spelling reform needs buy-in from the public exam system

In my view, the only approach that could stand any chance of success is to persuade education ministries and all universities to permit the revised spellings in all public exams, and for all school and university work. I don’t think it would be possible (at least for several years) to require only the new spellings, so the old spellings would have to continue to be permitted alongside the TSR spellings.

Focus TSR just on doubling and ‘magic e’

Although TSR is the least radical of the 6 systems shortlisted by the English Spelling Society, I see no chance of persuading education ministries to adopt TSR in its current form. The TSR changes are so radical that education ministries would be taking on too much risk if they were to require (or even permit) the use of TSR.

I would, therefore, cut down TSR to focus only on more consistent application of the doubling rule and the ‘magic e’ rule. Those 2 changes would:

  • affect the spelling of many words
  • rely on rules that already exist
  • apply those rules more consistently
  • be fairly easy to explain and understand

I believe that it would be easier to persuade education ministries to permit TSR if TSR is limited to those 2 changes. Those 2 changes would (in my view) improve English spelling enough to justify those changes, and would not be overwhelming.

Although many of the other changes proposed in TSR have some merits, they would simply cause too much resistance to stand any chance of being implemented. Those other changes could be left on the shelf as a package of further reform that could be considered in due course—though, realistically, not for several decades, because reform is costly and repeated rounds of reform breed exhaustion.

A specimen text

The paper introducing TSR provides the following text of the Gettysburg Address, spelled using TSR.

Gettysburg Address

Fourscor and seven years ago our faathers brought forth, on this continent, a new nation, conceeved in libberty and deddicated to the proposition that all men are creäted equal. Now we are engaged in a grait civil war, testing whether that nation, or any nation so conceeved, and so deddicated, can long endure.

We are met on a grait battle-feeld in that war. We have cum to deddicate a portion of that feeld, as a final resting place for those who here gave their lives, that that nation might liv. It is altogether fitting and propper that we should do this, but in a larger senss we cannot deddicate, we cannot consecrate, we cannot hallo this ground. The brave men, living and ded, who struggled here, have consecrated it far abuv our poor power to add or detract. The world will little note, nor long remember, what we say here, but it can nevver forget what they did here.

It is for us the living, rather, to be deddicated to the grait task remaining befor us that from these onored ded we take increassed devotion to that cause for which they gave the last fuul mesure of devotion – that we here highly resolv that these ded shall not have died in vain, that this nation, under God, shall have a new birth of freedom, and that guvernment of the peeple, by the peeple, for the peeple, shall not perrish from the erth.


  1. What a waste of time, and based only on RP pronunciation. “Cuvver”? What about the many areas where it is pronounced “coover” (with a short “oo”)?

    “bloo”? When I was at primary school in Chester, and we had elocution lessons (no, really), my form mistress, Miss Sadler, insisted that we should pronounce “blue” and “glue” as “blyoo” and “glyoo”, and “suit” as “syoot”. All right, that was in 1954.

    I wish the German spelling reform of a few years ago had never been implemented. They didn’t get rid of the annoying and superfluous “ß” (the “sharp s”) character. The sensible Swiss abolished it long ago. I can never remember the rules for using it, and it serves no purpose whatsoever.

    Those who have trouble with the current English spelling would have just as much trouble with the proposed new spelling. I am happily confident that it will never be adopted.

    1. Thanks for the comments, Alan. Actually, it’s explicitly not based just on RP pronunciation. The idea is that it is no more (and no less) based on a single pronunciation than existing spelling. There is a discussion in the underlying paper about how TSR tries to accommodate both RP and ‘General American’. Speakers of both varieties would continue to keep their own pronunciation and would be able to map from the spelling to the pronunciation not less easily than they can now. But, it is, of course, impossible to accommodate every variety of English pronunciation to an equal extent.

      I tried to summarise this part of the discussion in my section “TSR in different varieties of English”.
      The proposal does suggest introducing separate spellings for a handful of cases in which the RP and General American pronunciations of a vowel differ in ways that aren’t predictable. The examples they give are tomato, vase and buoy. I’m not sure the words concerned are numerous enough (or frequent enough) to make it worth doing this.

    2. I was surprised when I discovered that the German “ß” used to be used sometimes in English as well.
      The recentish German spelling reform did change the rules for when this character is now used. I did check the new rules a few months back, but can no longer remember them. I learnt the old rules just by practice and still have them internalised, without ever learning them explicitly.

Leave a comment

Your email address will not be published. Required fields are marked *