double Ls

You can buy this
I've just come from Scrabble club, where I often have to suffer the indignity of people putting down completely silly misspellings of words and saying "Maybe the Americans spell it like that." (The Association of British Scrabble Players started playing to the international Scrabble dictionary, called SOWPODS in American Scrabble circles, in 2001. The American National Scrabble Association has taken an isolationist position, and now only North America doesn't play the combined dictionary.) I take great joy in telling the people who guess silly "American" words that the American spelling system is more regular than the British. Unfortunately, no one pays attention to me when I do so. So, I'll subject this audience to my rant, since I can't know for sure if you're ignoring me or not, and can pretend that you're all fascinated by my opinions on spelling-rule complexity.

The American rule is: if the stress is on the syllable that attaches to the suffix, then you double any final consonant that follows a short (lax) vowel. But if the stress is elsewhere, you don't double the final consonant.

comMIT --> committing (not *commiting)
reFER --> referred (not *refered)
BUS --> bussed (not *bused)

EDit --> editing (not *editting)
LAbel --> labeling (preferred over labelling)
aBANdon --> abandoning (not *abandonning)
FOcus --> focused (preferred over focussed)
SEver --> severed (not *severred)

British English follows the same rule, EXCEPT when it comes to words ending in L, which are doubled after short vowels regardless of the stress. So, for no particular reason labelling, travelling and gambolling have one more L in the preferred BrE spelling than in the preferred AmE spelling. Notably the 1990 Concise Oxford Dictionary (COD) doesn't even mention the possibility of labeled, labeling and labeler. The American Heritage Dictionary lists the double-L versions after the single-L ones.

For some reason, focus is also an exception. While the COD and Microsoft's UK English spellchecker prefer focuses over focusses, I have been "corrected" for using only one s.

Ones that really strike my American eye as wrong are BrE dialling and fuelling. Since the l is preceded by a 'long' vowel (the diphthongs /aj/ and /ju/) in my pronunciation), they shouldn't have doubled consonants, just as one doesn't double the L in tailing or healing. They seem to come under the 'doubling' rule because dial and fuel are perceived as having two syllables each, with the latter one being unstressed--i.e. di-al and fu-el. The COD presents the BrE pronunciation as /dai(ə)l/ and /'fju:əl/--so definitely two syllables in fuel but not necessarily in dial. I'm not convinced that the second syllable in fuel is regularly pronounced. Better Half pronounces fuel with one syllable and dial with close to two.

I'm not a big one for spelling reform, so I don't mind that the two spelling systems differ. Just don't insinuate that American spelling makes less sense than British!


  1. I've learned something about American English here. I didn't know the rule about doubling consonants...I guess I had just memorized which to double and which not to double.

    I, for one, AM fascinated by your opinions on spelling-rule complexity, by the way! ;-)

    One more thing...when I saw the title of your entry today, I thought it might have something to do with the Welsh language. A Welsh colleague was JUST telling me yesterday that because his middle name is "Lloyd", strictly speaking his initials SHOULD be shown as "GLLW" rather than "GLW". I must admit that I know almost nothing about Welsh myself.



  2. Well, since you come from the US where the spelling system makes sense, I bet that you didn't have to memori{s/z}e the forms. You could have internali{z/s}ed the rule without reali{s/z}ing it, just as we do for the rules of English grammar (like where you can put an adverb and where you can't) that we don't know we know.

  3. I don't think I ever realised there was a dispute, except in dialling. Oh, and maybe focussed, since I always type it with two Ls and Word corrects it for me (so very irritating, when your computer thinks it know better...)

    My Northern accent does say 'fu-el' with two syllables, btw :)

  4. I thought of another one - I'm a journaLLer and I do journaLLing :)

  5. As opposed to a journalist, which again proves that AmE spelling is more regular. If BrE is going to double Ls before suffixes, why does it double the L in journaller and labeller but not in *journallist?

    American spelling rules ok! :)

  6. We don't make much sense, really, do we? I think it must be awful to have to learn English, it must be so complicated.

  7. Except that many Americans (even in dictionaries) do say 'bused'. Ick.

    And Rebecca? You can make your spell-checker learn your preferred spelling. Triumph over the machine!

  8. Bused is a funny one. I reali{s/z}ed that as I wrote it. I think that the one-s spelling has come about because bussed is homographic for the past tense of bus or buss. But there are few cases in which there will be an unresolvable ambiguity between the two busseds.

    1. Bused was the worst possible example: it's the only one-syllable verb that *doesn't* follow this rule in AmE. Major prescriptive style guides such as the AP and Garner allow only bused, always giving the reason as the collision with the other verb buss. Bused has a strong lead over bussed in American news media, even the New Yorker, where they usually want as many double consonants as possible.

      So bused is certainly not starred in AmE, quite the reverse. You may have gotten your impression from British media, where bussed has been much more successful. You could probably get a US/UK difference post out of this.

      Meanwhile, the plural noun the buses is far ahead of the busses on both sides of the Atlantic, dashing Fowler's hope in Modern English Usage (1926): "The plural of bus is usually buses ; this irregularity is explained by the fact that buses is still regarded as an abbreviation of the regular omnibuses ; when that is forgotten (& bus is now more usual than ’bus), doubtless buses will become, as it should, busses."

      I don't know why the noun would behave differently from the verb.

  9. I didn't even know the word buss until just now, when I looked it up and learned that it means "to kiss." Huh... thanks!

  10. And, yes, in Welsh LL is considered a single "letter". So are CH DD PH RH TH NG, the last of which follows G in the alphabet (not N), which can make looking words up a bit tricky if you forget. (Not that words start with NG, unless mutated, but it's still a long way between G and N.

    So in Welsh GLLW is right.

  11. But Br E Instalment v Am E Installment?

    Also Br E Trialist v Am E Triallist.

    If so, why?

  12. As far as I can tell, BrE instalment because the verb in BrE is spelt instal (though Oxford prefers install now). Similarly with BrE fulfil/fulfilment versus AmE fulfill/fulfillment. However installing and fulfilling are spelt with two Ls in BrE (like AmE), following the rules discussed in the post. It stays one L before the ment suffix because the L is only doubled before a vowel. (See this later post's comments for more on the 'rules' for removing Ls before consonant-initial suffixes.

    Why does BrE spell some verbs with one L, while AmE spells them with two? No idea. The double-L version has more connection to the origins of the verbs: Old English fullfyllan and Latin installar. Instil is another example. Notice that similar one-syllable verbs are spelt with two Ls in BrE, like AmE: full, stall, still.

    As for triallist, I can find no evidence of this in AmE, so you'd have to show me some examples to convince me. Where I've found AmE examples, they refer to lists of trials not people who take part in trials.

    My Concise Oxford gives no double-L variation for trialist, though COD does include other US single/double-L variations. I'm not finding it in US dictionaries. Chambers, makers of Official Scrabble Words lists both. Neither spelling is listed as American, and neither is in the American Official Scrabble Players Dictionary.

    Why does trialist break the mo(u)ld regarding BrE double L's? It's unlike the other examples in that it's not derived from a verb. (And to be technical, it involves a derivational, rather than inflectional suffix--see here for definitions.) It's the noun trial plus -ist.

  13. I've been learning English over 6 years now and the moment I started to realize the differences between AmE & its British counterpart was when I went to college. Now I personally use AmE (and know the rules about spelling and other aspects of the "two languages"), but a few years ago if I had been asked which one I used I'd propably have had no idea... :I
    btw does anybody know any interesting websites on American Pronunciation??

  14. Everything else pales before the random use of -our and -or on the one hand, and -er and -re on the other. Particularly in derived words: why honour and honourable but honorary, with occasional outbreaks of honourary?

  15. What about "format"? I emphasize the first syllable, yet it seems "formatted, formatting, and formatter" are correct...

  16. The problem that I find with the single L in words such as 'canceled' and 'traveled' is that when I attended primary school not so many years ago I was always taught the use of the 'magic E', (i.e. the 'a' in 'rat' changing its pronunciation when an 'e' is added at the end to make it 'rate' etc). My mind applies this to 'canceled' to the point where to me it looks like it should be pronounced 'canceeled'. A bit of a strange comment I realise but i wonder if this applies to anybody else?
    I also believe that BrE tries to be more standardised in some respects with the LL ending being applied most commonly. AmE distinguishes between the L and LL with words such as 'excelling' and 'signaling' whereas BrE would use the LL for both. I wonder if the difference between the two is necessary to be pointed out since a 'standard' of LL in BrE would appear to be more simple. I'm not saying that I believe AmE to be wrong in any sense, variety is the spice of life after all, but it is just a thought...

  17. That's funny Sean, I find myself wanting to pronounce cancelled can-celled because I know the rule is supposed to be based on syllable stressing, like lynneguist said.

  18. AmE also has its exceptions, if you look hard enough e.g. "originally".

  19. That's not an exception, Brendan, it's a different set of rules. In the rule we were discussing, it's whether or not to add an 'l'. Looking at original+ly, there is an 'l' in the suffix, and thus there are two 'l's not because of an 'l'-adding rule, but because there's no such thing as an 'l'-deleting rule.

    Also note that the ones we've discussed here are all verbs. So, originally falls outside the discussion. Sorry if that wasn't clear in the post!

  20. It seems to me that even for speakers for "dial" and "fuel" have one long vowel, "dialling" and "fuelling" are always "di-alling" and "fu-elling". Am I mistaken?

    I think the British rules are based on the resulting words themselves, not on how the root verbs were :)

    [Thus, as journaller and journalist are pronounced differently... Of course, I'm not from a Anglophone country so it's possible that what I hear is spelling pronunciation :)]

  21. for me, there's never an 'e' in fuel(l)ing: it's /fjuliN/.

    1. I pronoun e fuel as few-al, and dial as dye-al. British Canadian.

  22. It strikes me none of these rules were invented by people with dyslexia.
    To all scrabble players:
    I hate standardised spellings

    Maybe it's all right for the purpose of a game; like chess or something with emphasis on logic and obscure rules, but in real life, if it works (if it's possible to understand what it means) it should be allowed.
    Except, possibly, in the case of homophones and things like that.

  23. Some dyslexics deal as well with irregular spellings as with regular, because they can't sound things out and therefore have to memori{s/z}e spellings, while others can't memori{s/z}e spellings and need phonetic renderings. The former type is nearly non-existent for Chinese, the latter type for Spanish. English gets the worst of it because our spelling system has elements of regularity and irregularity. The term 'dyslexia' is a cover-all term that doesn't distinguish among the many ways of reading that are 'dyslexic'.

    But, still, there's a problem with the 'you should be able to spell it any way you like' strategy in that what works for you won't necessarily work for the person who's reading your work. So, if an American were writing to a BrE speaker (or even a Texan to a New Yorker), the vowels they might tend to use when attempting a 'spell it like it sounds' system would probably be very different. I know that I've had to think much harder about 'phonetic' spellings from BrE speakers than from AmE speakers--in particular because of what they do with 'r's--with BrE speakers often using them in places where they want to indicate a long vowel, rather than an 'r' sound. (But that's a topic that's come up here a few times.)

    The nice thing about standardi{s/z}ation is that people are using the same system, no matter how they pronounce the words, and so we can all communicate. What's more, we can read work written by people from different times when things were pronounced differently. There was a big reform of the writing system in Japan in the 20th century, and people who have grown up in the new system report that they can't read the letters their grandparents wrote. How sad!

  24. I came here searching this blog for double Ls, but I was thinking of another context: words like skil(l)ful and wil(l)ful. It seems that BrE drops the l, while AmE keeps it — but would you know if this rule is recent? I'm reading an old American book (published in 1925) which uses AmE spellings everywhere (honor, etc.) but also uses wilful and skilful (also, fulfilment). I wonder if spelling rules changed in AmE since then?

  25. oh man. i laughed so hard at this, thanks for it.

    how about Fourty and Forty? if so, then why not Forteen? (it's Fourteen isn't it?)

  26. OK, perhaps the spelled: can-suh-LAY-shun. Then cancellation makes sence. Though, still formatted and CANcellate.

  27. I agree with Sean, if I see "canceled" I want to pronounce it so that it would be almost the same as "concealed" (except for the first vowel, obviously).

    And while a rule might be more rigorously applied in AE, I'm not sure that necessarily helps with spelling. When teaching in America I saw a lot of kids using single l's when a double is required even in AE, e.g. "rebeled". A mistake you'd never get in BE (though "editted", etc are more common there). So it "making more sense" if you look at syllable stress and vowel length in various contexts seems irrelevant when everyday users of the written language are making effectively the same mistakes (just reversed) with both systems.

  28. This British former physics teacher and one-time photographic judge finds 'focussed' and 'focussing' sets her teeth on edge.

  29. I don't see what's so simple about single L.

    We Brits double our consonants before a vowel without having to think about it. Stress isn't a consideration.

    The downside is that we have to learn as exceptions such spellings as edited, abandoned, severed.

    Both systems have regularities and exceptions.

    As I've said on another thread, the argument that traveling signals a rhyme with stealing is just a joke. All the words with that rhyme are spelling with ealing or eeling. But we do have woods with spellings like gambol(l)ing where the -oling spelling signals a long o. There's poling and doling and, no doubt, others.

    Of course I'm not saying that British spelling is easier or more logical in every way. But it is easy and logical in a couple of ways:
    • We don't have to pay any attention to stress. Why is this a good thing? It means we don't have to start by recognising pronunciation when we read. It means we don't have to analyse the stress of the word when we write.
    • It doesn't conflict with the rule that identifies a single consonant as a pointer to a preceding long vowel.

    The Webster reform is fine for words which the reader knows and recognises. But if the word is unfamiliar, or if it doesn't immediately spring to the reader's inner ear, then stress is still to be discovered.

    If I fail to recognise the word and piece together the spoken word fragments TRA-VELL-ER, I'm pretty certain to recognise travel(l)er. But If I piece together TRA-VEEL-ER then it might take a little longer.

  30. Two important problems with English spelling arising out of piecemeal attempts at consistency in the past:

    1. The vowel letters A, E, I (with variant Y), O, U may represent five pairs of sounds, which
    once upon a time were long and short pairings
    but now represent ten different sounds with no obvious similarity within each pair.

    2. The six letters A, E, I (with variant Y), O, U, Y may also represent different sounds again when they're i unstressed syllables. This is especially true of E.

    British spelling convention addresses Problem 1 and ignores Problem 2. The Webster reform (US spelling) addressed primarily Problem 2.

    British spelling isn't 'right', but it isn't dumb either. It just has a different priority.

    The value of the British priority seems diminished by the special case of E — and further diminished when the consonant letter is L. But for the other vowels there are examples where a single consonant usefully signals a long vowel.
    • For O and I
    —there are contrasting pairs such as holy~holly~wily, Willy
    • For A and U, despite conventions for different sounds represented by all and ull spellings,
    — the convention remains that they represent the long variants before single consonant as in whaling, ruling
    — the convention that they represent the short variants before double consonants does operate in part, helping us to recognise such words as rally, gully

    Even letter E provides some examples where the British convention helps to recognise words.
    • The short e before double consonants is widely used as in smelling. Indeed, the convention is usually extended to word-final as in smell. Still, there's an approximate pair in gel~jelly.
    • The long e convention is occasionally useful. English speakers in countries without a place called Ely or who haven't heard of Sir Peter Lely

    We British as school children explicitly learn or implicitly absorb a simple rule which helps us to recognise a great many many words spelled with A, I/Y, O, U followed by L or LL. It's less useful for recognising words spelled with E followed by L or LL, but it does no had and is occasionally helpful. The merits of the rule are:
    1. It's extremely simple.
    2. You don't have to think about it

    Like all simple rules, it will let you down some of the time. But we think the merits outweigh the more intelligent Webster rule which involves a bit of thought.

  31. David Crystal has some interesting thoughts on this. It's among the archive of old publications he's putting on line.

    Clicl here, then click the dowline button under Single and double 'l'.

    Echoing Pam Peters, he cites cases where British spellings are inconsistent and at least one American anomaly: crystallized. There are also words with variable spelling — examples given are medal(l)ists, panel(l)ling, towel(l)ling, bejewel(l)ed, carol(l)ling — which Pam Peters (writing from Australia) proposes should be regularised internationally with single L.

  32. If I were to speak "more properly", then I would say dial and fuel as two distinctive syllables (di-ull, rather than dile, for example) and fuel obviously, if only from a poetic sense, did have more than one syllable - see "gathering winter fu-oo-el).

    Granted, in my normal way of speaking, I would slur those two words into one syllable each.

  33. In which cases are double Ls such as "modelling" acceptable in AmE and why? I edit texts regularly for a client who wants AmE, unlike most editing clients in Finland who want BrE, and I am always a little stumped when he writes "modelling" and the American spell checker doesn't mark it as wrong. I know the New Yorker uses it, but is it common otherwise in AmE? Should I change it to "modeling" if he wants a more "standard" AmE, whatever that means? I imagine my client has just learnt BrE in school and doesn't know about some of the finer spelling differences, such as single versus double Ls.

    1. Hi Ian, AmE is less fixed about these spellings than BrE is. At school I learned that both were OK, and both were listed in my school dictionaries (which are more selective than adult dictionaries about what they let in). It's kind of like the ise/ize difference in BrE, where the tradition has been varied and gets more solidified over time. On 'mode*ing', the GloWBE corpus has over 600 hits for two Ls in AmE, but over 3000 for two Ls. When I'm editing for AmE, I do reduce the Ls.

      The main thing, of course, is just to be consistent in a document.


The book!

View by topic



AmE = American English
BrE = British English
OED = Oxford English Dictionary (online)