Saturday, 2 March 2013


Stibium looks like this:


Stibium /'stɪbiəm/ is the Latin name for the chemical element antimony. Don't worry, this is not a blog post on chemistry, but rather one about the stress patterns of this word. 

Triumphal Chariot of Antimony by the Benedictine monk Basilius Valentinus
The antimony sulphite has been known for a very long time. In the 17th c. the Benedictine monk Basil Valentine published a monograph on the chemistry of the metal (see here for a much more detailed description of the etymology and history of antimony).

But I'm interested in the stress pattern of antimony.
Samuel Johnson in his Dictionary of the English Language (5th ed. 1785) assigns initial stress to the word:

The same stress assignment is found in Thomas Sheridan's Complete Dictionary of the English Language (3rd ed. 1790):

In John Walker's Critical Pronouncing Dictionary ... of 1824 the word is also marked as initially stressed:

Noah Webster in his American Dictionary of the English Language (4th ed. 1840) likewise records initial-stressed antimony:

Addendum:  The first edition of the OED records the initial stress pattern of antimony only:

In their Phonetic Dictionary of the English Language of 1913 H. Michaelis and D. Jones indicate the pronunciation of antimony as /ˈæntiməni/ - no change of the stress pattern:
Michaelis-Jones 1913
I then checked the 1st ed. of 1919, the 9th of 1958 and the 11th of 1964 (the latter is a reprint of the same edition of 1963 "with corrections and minor revisions by A. C. Gimson" (p. vi)) of the English Pronouncing Dictionary. In all three editions the word has initial stress. I don't have the 13th edition at hand for which Gimson took over the editorial baton; but a friend of mine told me the entry looks exactly like that of the 12th edition. Even the 14th ed. of 1977 (with Gimson still being the editor) leaves the stress pattern unaltered.

From the 15th edition onwards it's Peter Roach, James Hartman and Jane Setter who've taken responsibility for the (C)EPD. And it is from this edition onward that we are offered an additional, second pronunciation variant - the one with antepenultimate stress:

Is there any meaning behind this particular sequence of pronunciations? Here is what the editors write:
Where more than one stress pattern is possible, the preferred pronunciation is given first and then alternatives are listed. (xv)
As an addition to the prefix anti- we find the following note in CEPD18 (p. 23):
Note: Prefix. Numerous compounds may be formed by prefixing anti- to other words. Most often, these compounds carry primary or secondary stress on the first syllable, e.g. antihero /ˈæn.tiˌhɪə.rəʊ/ /-t̬iˌhɪr.oʊ/, anti-icer /ˌæn.tiˈaɪ.sər/ /-t̬iˈaɪ.sɚ/, but there are also cases in which the second syllable takes the primary stress, e.g. antinomy /ænˈtɪn.ə.mi/.
This description causes me tummy ache for two reasons:
1. How does the dictionary user know if a formation containing the constituent is a compound or not?
2. Prefixes are bound morphemes and compounds are combinations of at least two free morphemes. I would never ever classify antinomy as a compound. Antihero and anti-icer are neo-classical compounds with anti being a combining form.

LPD in all its three editions has initial-stressed antimony only.

PS: Other words ending in -mony are acrimony, alimony, agrimony, ceremony, hegemony, matrimony, palimony, parsimony, patrimony and testimony. A lot of -mony.


  1. Now give matrimony the same treatment?

  2. Well done Petr. I don't see antimony as a compound either. Johnson's "antimonk" appears to be a popular or folk etymology, obtained by spuriously deriving morphemes where there are none (ignoring for the moment your observation that anti is a prefix and not a free morpheme anyway). And I would imagine that Johnson was retelling an established tale. However, anti is used freely nowadays (and your spelling checker doesn't object to it), for example "he's anti everything". So it's not surprising if people have started splitting antimony into two free morphemes and letting that modify their pronunciation.

    1. It occurs to me too late that analysing anti- as a morpheme, whether falsely or correctly, should put the accent on the following morpheme: anti'monk, anti'war, anti'drugs etc. This would lead to *anti'mony (like anti'money), leaving an'timony still unexplained. Perhaps it's analagous to the ki'lometre vs 'kilometre dispute (while mi'crometer, ther'mometer, ga'someter etc are OK).

    2. There seems to be a difference between measuring instruments, stressed on the antepenultimate, and units, which at least tend to have the stress on the first part. 'Micrometre (µm) vs mi'crometer (₱).

    3. Sidney: If you accept {anti-} as a morpheme, why should this shift the stress to the following morpheme? There a re quite a few such formations with primary stress on {anti-}, e.g. antihero, antibody, Antichrist, antidote.

    4. There seems to be a difference between anti-'X = somebody or something that is against X and 'anti-X = the opposite or adversary of X on basically the same level. (The meanings might overlap, but secondarily only, I'd say. I suppose the Antichrist is anti-Christ, but what do I know about theology?)

    5. When anti is used attributively, it follows the usual rule of accenting the end of the phrase - like blue FLOWERS, churchWARDEN, anti-WAR (ignore for now writing conventions like joining with a space, a hyphen or nothing, that's a different issue). Over time, many of these constructions have become true compound nouns (one single noun) with the first element accented, and I suppose ANTIchrist is an example. This is an on-going process, look through Daniel Jones' dictionary for many examples like churchYARD (now CHURCHyard). Typically, you hear both versions during the transition. I'm not sure about antihero. It could be still moving. And so on.

    6. Sidney, you don't think there's something to the semantic difference I suggested?

    7. Off the cuff, it looks like there could be. I can't think of enough different meters on the spur of the moment. The instrument you quote has o which belongs to the first element - micro, like thermo (ther'mometer), pyro (pyro'meter), photo- (pho'tometer). The gasometer I mentioned has a spurious o, gas+o+meter. Being British I'm thinking of the giant cylinders at the gasworks for storing coal gas, and for measuring the stock by noting it expanding (and Larousse has the same for gazomètre, so that does seem to explain that o). My American dictionary has a different meaning, a laboratory instrument for measuring gases, loaned from French gazomètre. It's too late at night to start off on a trail of French word formation. Otherwise linguistics have always had an escape route from cases that fall out of line, by referring to analogy. One such is kilometer, ought to be 'kilometer fitting your distinction, but is commonly pronounced ki'lometer. Then there are other measuring instruments like 'voltmeter.

    8. Escape route sounds like an excuse, while those cases are indeed easily explained - ki'lometer is clearly the younger pronunciation, maybe still considered somewhat uneducated, and there are enough people around who still say 'kilometer without any agenda (I do, for what it's worth). Voltmeter doesn't follow the word formation pattern anyway.

    9. As I said, it looks like you're correct about the -ometer words, where the first element -o seems to be of greek origin, another example is ta'chometer. There are more recent creations like ga'someter (that might be loaned from French). Spee'dometer and mi'leometer have just occurred to me, and they look purely English in origin. Other recent meters seem to follow the other pattern, like 'voltmeter, 'altimeter. For some reason, the word kilometre has been treated differently, the unit of distance being accented 'kilometre by some and ki'lometre by others. And as long as the UK and elsewhere still measures distance in miles they won't need a meter to measure kilometres. I don't know what the Irish have on their dashboards since Eire went fully metric. For the record, I've always said ki'lometre, I'm 80, and I have a PhD. No offence intended, "uneducated" pronunciation has never been a function of qualifications. I certainly don't claim to be the first person to say ki'lometre, so that means we have to go back at least one generation, then we're nearly into the 19th c. Daniel Jones listed it in the 1957 edition of his pronouncing dictionary.

    10. Please don't get me wrong - I don't consider ki'lometer "uneducated" at all myself, and today, if anything, 'kilometer is marked. The OED, not updated, says (ˈkɪləmiːtə(r)) Also with pronunc. (kɪˈlɒmɪtə(r)), prob. under the influence of such words as speedometer, thermometer. […] (The stress is marked by Webster (1828), Craig, and Cassell as kiˈlometre.)]

    11. Webster (1828) is a good find. I was trying to figure out when Brits would want to talk about kilometres, and I had a hard time getting beyond school maths, and the odd book that might happen to mention them. Travellers. WW1 and the west front. But established by 1828, that takes as back to Waterloo at least. The metric system was createded in France in the 1790s, and introduced in 1799, which could mean that both pronunciations were there from the beginning, being so close to 1828. Wikipedia gives the metric system a prehistory among mathematicians back to 16c.

  3. Peter Roach and colleagues were spot-on in recording the British subvariant stressing an`timony from 1997. The word doesnt exac·ly figure in ev·rybody's ev·ryday conversation so it's not surprising that it didnt apprear in 2008 in LPD3. My own first record of hearing it was in a radio medical specialists' discussion of 2004 and I've noted it since from the distinguisht physicist Jim Al Khalili. It's rather puzzling to think how it cou·d've come about unless the rather abstruse logician's word an`tinomy has been perceived as a better model in some way.
    An item I commend to Kraut for inclusion in his historical collection is that Murray in 1885 in NED (= OED1) gave only the stressing `antimony.

  4. If antinomy is abstruse then so is abstruse itself. In other words: What makes you think that it is abstruse?

  5. K, how did you make those 'screenshots' from LPD and CPD? You took a photo of the entry in the book and then cropped it? Is there any other way?