World's Fair


Science scout twitter feed

The reason why there were two ways of saying the element aluminum/aluminium has always been one of the those things that made me go “hmmm” But by the same token, it’s also always been one of those things that never stuck around in my consciousness long enough for me to look it up.

Well, lucky for all us, Michael Quinion over at World Wide Words does an awesome job of going into the lexicon of these words, paying particular attention to why two forms exist – specifically, why the Brits say “Alumininium” and why Americans say “Aluminum.”

It’s actually quite intriguing how it came to be – first starting with Sir Humphry Davy’s not being able to make a decision:

Sir Humphry made a bit of a mess of naming this new element, at first spelling it alumium (this was in 1807) then changing it to aluminum, and finally settling on aluminium in 1812. His classically educated scientific colleagues preferred aluminium right from the start, because it had more of a classical ring, and chimed harmoniously with many other elements whose names ended in -ium, like potassium, sodium, and magnesium, all of which had been named by Davy.

And why it stuck in the “aluminum” form within the United States, had more to do with publications like the Noah Webster’s Dictionary which in 1828 listed only the -um form in its pages. Such preferences crops up here and there during this time in history and also included the 1913 edition of the Unabridged version.

In any event, it was regarding this period of time, that Quinion writes:

It’s clear that the shift in the USA from -ium to -um took place progressively over a period starting in about 1895, when the metal began to be widely available and the word started to be needed in popular writing. It is easy to imagine journalists turning for confirmation to Webster’s Dictionary, still the most influential work at that time, and adopting its spelling. The official change in the US to the -um spelling happened quite late: the American Chemical Society only adopted it in 1925, though this was clearly in response to the popular shift that had already taken place. The International Union of Pure and Applied Chemistry (IUPAC) officially standardised on aluminium in 1990, though this has done nothing, of course, to change the way people in the US spell it for day to day purposes.

Anyway, there you have it- looks to me like it was all Webster’s fault.


  1. #1 Don in Rochester MN
    April 8, 2009

    Uhhhh . . . that’s MICHAEL Quinion! (I’m a big fan of his . . . and, therefore, somewhat of a nit-picker.)

  2. #2 David Ng
    April 8, 2009

    Oops. You’re right, and apologies to Mr. Quinion for that – how did I get Milton from Michael? Hmmm….maybe I looked him up in Webster’s…

  3. #3 Flip van Tiel
    April 8, 2009

    There is more to worry about: there is the natrium/sodium pair (for element Na) and the kalium/potassium pair (for element K). How come? Does it matter?

    Much more serious (and in these critical monetary times possibly crucial) is the use of billion for 10^9 instead of 10^12. The latter (10^12) is the original ‘number’ and the word indicates that it is somehow a ‘bi’ of something, namely the square of a million. A logical consequence is that a trillion is the third power of a million or 10^18).

    The anglo-saxon billion (10^9) is in international parlance elsewhere a milliard (originally meaning ‘grand million’). Examples: light travels approximately 9.5 billion km annually and the world’s population at present is approximately 7 milliard. The problem with calling 10^9 a billion, 10^12 a trillion, and 10^15 a quadrillion, is that the prefixes bi-, tri-, quadri-, etc. make no sense in numerical terminology anymore. For instance, what is 10^9 two (2) of, or what is 10^15 four (4) of?

    I brought this up at the 2008 annual conference of the Academia Europea last summer when cosmologist Lord Martin Rees, the president of the Royal Society had presented the mostly continental European audience with a fascinating story about events that, he said, had taken place billions of years ago and were now billions of light years away, al along meaning milliards of years and milliard of light years. Lord Rees was kind enough to acknowledge that he had probably caused some confusion among the audience, but that he felt the milliard had eventually lost the battle.
    That was not received as a valid reply, as I found out later, talking to several of the delegates, but it will indeed be difficult to turn the tide… unless perhaps the civilized world could decide on using, say, million, 1000-million, billion, 1000-billion, etc. for 10^6, 10^9, 10^12, 10^15 etc.

    I rest my case, but not without concluding that we clearly have here an illogical left-over of the Imperial System of units and measures.

  4. #4 Sam C
    April 9, 2009

    Flip van Tiel is worrying unnecessarily I think.

    Firstly, billion at 10^9 and trillion as 10^12 is fairly universally accepted in English-speaking countries nowadays. Certainly it’s what we mean in the UK even though we certainly had billion as 10^12 before the 1960s.

    Secondly, how frequently can there realistically be confusion? OK, a layman told the universe is 16 billion years old might not know which is which, but does that matter – it’s 16 very-long-times either way!

    Thirdly, we don’t use trillions much at all. Billions are useful for populations and large amounts of money – sometimes trillions are used, but I think many people will mentally shift them back to billions for comparability. We don’t really have trillions of anything! Or, when we do, it’s bacteria or atoms or grains of sand, exponent form is more useful. And again, the number means just “a heck of a lot” when it’s quoted!

    Fourthly, evidence for it not causing confusion: most European countries use a comma to separate decimal fractions and points to separate grouped thousands while English-speaking countries do the opposite. When Europeans document calculations in English, there is sometimes a random mix of commas and points for separating decimals, but there has never been any problem due this confusion in the engineering world: people know whether 1.234 is over a thousand or nearly one.

    Your suggestion of changing the meaning to billion=10^12 is one trillion per cent idiotic though! You could make an argument for new names maybe, but changing existing ones to mean something different? We’ve undone the confusion now, it’s only a few people (like you?) who think it’s still confusing, so you propose to bugger it up for everybody else. Thanks! What next, rename the digits 0=eight, 1=five, 2=four, 3=nine, 4=one, 5=seven, 6=six, 7=three, 8=two, 9=zero because it would be so much better if they were in alphabetical order?!

    On aluminium, I think it doesn’t matter. They’re so close that there’s easy mutual comprehension, and a little bit of variety is nice, isn’t it?!

  5. #5 Michael Quinion
    April 11, 2009

    Many thanks for correcting my name (Milton Quinion might be an obscure American cousin of mine, but I’ve never met him). But while we’re correcting typos, would you change the name of the site to “World Wide Words”?

  6. #6 David Ng
    April 11, 2009

    Hi Michael,

    Apologies again. Changed the typo and thanks for dropping by. Again, it was an excellent piece, and I’m really enjoying your site lately.

  7. #7 Flip van Tiel
    April 11, 2009

    Re @4 by Sam C

    The content and especially the tone of the commentator are counterproductive. I merely provided an example of a potential source of (at least occasional) confusion among not “only a few” people but many millions who happen to express themselves primarily in other languages than English. Apart from that the fact that the layman may have a number sense that goes only a bit further than the one-two-many of the turtle and the seagull (see Stanislas Dehaene for more on the human ‘number sense’), the demographic fact raised by Sam C is irrelevant to my argument. What I had in mind is a problem that concerns those who have to deal professionally with very large and very small numbers and quantities, that is, it concerns most scientists. That this is indeed the case, the commentator may be able to comprehend by consulting where a variety of scales between 10^-24 and 10^24 will be found. By checking what the various prefixes mean, he will also find out that the ISU has taken great care in choosing terms that have an intrinsic numerical meaning [derived from a number of languages other than English, e.g. femto- (Swedish fem=5 and an -o for negative powers, for 1000^-5 = 10^-15) or exa-for (10^3)^6 (Greek (h)exa=6)]. Using 10^3 as the basis for numerical prefixes is a great idea. It may, however, take a short while to introduce this for everyone, including the commentator, but perhaps not longer than it took to officially adopt the 24-hour clock and the metric system (at least in the UK).

    There is no merit in a further discussion and therefore I shall not continue on this theme.

    PS – One final point on the comma/point issue raised by Sam C. Nobody that I know has ever made an issue of this. It became problematic only when we started using text processors for writing and spread sheets for calculations.

  8. #8 Vicki Donovan
    March 28, 2010

    The Brits don’t say ‘alumininium’, we say ‘aluminium’! 😉

  9. #9 thomas sabo ringe
    July 15, 2010

    Your suggestion of changing the meaning to billion=10^12 is one trillion per cent idiotic though! You could make an argument for new names maybe, but changing existing ones to mean something different? We’ve undone the confusion now, it’s only a few people (like you?) who think it’s still confusing, so you propose to bugger it up for everybody else. Thanks! What next, rename the digits 0=eight, 1=five, 2=four, 3=nine, 4=one, 5=seven, 6=six, 7=three, 8=two, 9=zero because it would be so much better if they were in alphabetical order?!

  10. #10 Will Killyou
    Just ahead of you
    September 3, 2012

    The English language has dominated more than any other due to the colonial and subsequent commercial success of Anglican nations and territories. For this reason the English language in it’s various forms (uhem) has become ubiquitous. Numerical naming conventions should NOT be changed from Anglican convention. The world will simply have to adjust to the dominant culture; no tolerance should be allowed for the whiny, the wimpy and the wussy! The ISU is blasphemous in it’s “rulings” and in turn should be disbanded immediately!

  11. #11 Jheri Cravens
    September 9, 2012

    It’s, that is, eye-tee-apostrophe-ess, means, always and forever, IT IS. If you wish to suggest that something belongs to it, you must say, “its.” I understand that this is contrary to the rules for all OTHER possessives. It is, or I might say, “it’s,” which means IT IS, nonetheless correct.

    You just wouldn’t believe how that overworked, MISUSED apostrophe grates on — and distracts — those of us who know better. Please join us.

  12. #12 David Norley
    October 7, 2012

    Hiya, I’m English, living in France. I’m amazed that no-one’s noticed another spelling mistake, look at your English version of the ‘word’ in your second paragraph, OMG!! You’re confusing everyone mate!! SPELLCHECK if you need it, works wonders, (hope that doesn’t offend), Thanks though, it was an interesting read, I’m helping a poor French friend that had the awful misfortune of learning English in America!! Trying to put him back on the right path! (Only kidding!) :-)

  13. #13 David Norley
    October 7, 2012

    Oops! I’ve just seen that…

    Vicki Donovan

    March 28, 2010, 10:24 am
    The Brits don’t say ‘alumininium’, we say ‘aluminium’!

    Sorry, I’ll shut up! Bye.

  14. #14 Jessica
    Hong Kong
    March 12, 2013

    I dont think there is a point to argue the way we call the billion trillion etc even though I can understand Flip’s annoyance and how it can be confusing. However, it is just a matter of which counting scale is adopted. Most western countries use the long or short scale (US, UK switched over):
    If you look at the table it makes sense. In the short scale the increments jump by the 1000 instead of 1,000,000.
    so one billion refers to 1000 x 1000^2 so there’s your (bi) =D

    If you want to get confusing you can go into asian counting scales where there’s a word for 10,000, 10^8, 10^12, 10^16, and 10^20.

  15. #15 Sonja Brignoni
    April 25, 2014

    great post I really loved it

  16. #16 Walker
    May 29, 2014

    Sorry Jessica, it doesn’t make sense. You are saying that 1000^3 is a Billion, (1000 x 1000^2) A billion is supposed to be a Million squared, not a million times a thousand. Otherwise, why not call 10 thousand a miilion as the next number set? Why even have 10 or 100 thousand?
    It has been made short form to appeal to the uneducated masses who don’t know better. Is America not a Trillion dollar economy? Is inflation slowing? The long form gave us somewhere to go that made sense in an almost infinite number of possibilities, the short form does not.

  17. #17 Dan Wylie-Sears
    July 5, 2014

    Are thousands the base of the naming sytem, or are you going to have sub-stacks of names all the way up? If you’re going to have thousand-millions before you go to billions, then you should have million-billions before you go to trillions. So if “billion” were going to be 10^12 instead of 10^9, then “trillion” would have to be 10^24.

    “Mill-” is thousand. “Million” is “big thousand”, i.e. thousand squared. “Billion” is big big thousand — two “bigs”, or the second step of the naming system. Trillion is big, big, big thousand — three bigs, or the third in the sequence of names. And so on.

  18. #18 Ryan
    November 2, 2014

    Walker, Jessica is right. Your, base is 1000, to get to billion one has to multiply their “base 1000” by two more 1000’s
    i.e. (base 1000 x 1000^2). For trillion, one multiples their “base 1000” by three more 1000’s. For quadrillion one multiplies their “base 1000” by four more thousands. And so on.

    Going back to the long scale as a standard is not worth the effort. The long scale goes up by 10^6 (example 10^6 x 10^6 = 10^12 ) so we have 10^6, 10^12, 10^18, and so on. All very nice, but what about the milliard, it’s 10^9, scaled by 10^3? The milliard doesn’t fit very nicely, in the long scale, especially when we’re are playing with base 10 and when we’re scaling by 10^6. The short scale shifts by 10^3 which fits even better for a base 10 (example 10^6 x 10^3 = 10^9, 10^9 x 10^3 = 10^12, and so on).

    By shifting to the short scale we have eliminated having to add new names to the long scale, such as milliard, for those numbers scaled by 10^3. We have also made the scale more accurate, relative to you data, by scaling with 10^3.

    Now looking back, the short scale is a much better and more practical scale, nothing has been added in ways of names and now we have a more accurate set. The only thing that has been changed is what we scale by (i.e. 10^3 instead of 10^6). In this case, the important thing to the scientist is accuracy and speed, all of which the short scale accomplishes by shifting with 10^3 instead of 10^6.

    As for the lay person, all they know is that billion means a lot, in fact it is very difficult to picture one billion even for a scientist.

    Going back to the long scale would be incredibly confusing. We would have to create new names for every scaling of 10^3. All data from the short scale would be have to converted and so on and so forth.

    In conclusion, Flip, the short scale will always dominate.

  19. #19 Cliff Bramlett
    United States
    July 3, 2015

    Just a quick note, David Ng: Mr. Quinion leaves out a possibly important detail regarding the word etymology.

    That link clearly quotes the authors as having changed the name without the input of the Sir Humphry. Whether we think he did a good job of naming an element or not, it was his discovery to name.

  20. #20 Christian Parker
    January 5, 2016

    Cliff is right. I’ve read this elsewhere as well as his source. Sir Humphry named it “alumium” then “aluminum.” It was other British that decided to go with “aluminium.”

  21. #21 David C.
    United States
    February 11, 2016

    Walker, the reason the short form works is not some low-brow appeal to the unwashed masses. It’s used for the same reasons that we use decimals, scientific notation, and have abandoned the Roman numeral system. It’s mathematical efficiency and not simplicity.

    In scientific notation, as soon as you see your exponent you know exactly where your math is taking place. There are no long descriptions of 10’s and 100’s and 1000’s and 10’s of thousands of anything. One hundred sixteen thousand two hundred twenty-nine million is mess. It’s a big number on top of a bigger number. It’s silly to say, and even less fun to add to a similar number.

    You work in units up to 999 of said units and then you move on. If I say 155.2 trillion, you know exactly where my math is. If we want to add 431 trillion to that, it is both pleasant and relative.