Spectrums
« previous post | next post »
Reading the comments on Sunday's post about verb agreement with data ("Scientist spotting",5/22/2022), I was reminded of a long-ago tussle about a different aspect of Latin morphology in English borrowings. What's the plural of spectrum? Is is "spectra" or "spectrums"?
In 1979, an editor of Scientific American asked me to write an article about progress and prospects of computer speech recognition, which was then in the middle of an "AI Winter". I enlisted my Bell Labs colleague Steve Levinson, and we submitted an article explaining the then-current technology, which was based on word- or phrase-level acoustic pattern-matching. This turned out not to be what the editors wanted, which apparently was a focus on expectation-driven ("top-down") methods.
Since (at least then) the magazine's ideology was that articles are written by scientists rather than journalists, they "fixed" the problem in what they called "copy editing". The result was a draft that had almost nothing in common with what we had submitted, not just in exposition but also in content. After several months of acrimonious tussle, they agreed to a compromise version, mostly written by us, which appeared in 1981 as "Speech Recognition by Computer".
The arguments about content (and about the magazine's pretense of scientists as authors) are not relevant here. But one surprisingly acrimonious disagreement was over whether the plural of spectrum should be "spectra" (what we wrote) or "spectrums" (what they wanted).
Since this was about 18 years before Google was founded, and 24 years before the launch of Google Scholar, we could only wave our hands in the direction of what we perceived to be the norms of our field.
Now I can ask Google Scholar about usage before 1980, resulting in an estimated ratio of about 68-to-1 in favor of "spectra" over "spectrums" (743,000 to 11,000). Current Google Scholar counts are about 24-to-1 in favor of "spectra" over "spectrums" (5,520,000 to 231,000), so apparently "spectrums" is gaining ground. And even in pre-1980 times, "spectrums" was used more commonly than I perceived at the time.
In any case, the Scientific American editors eventually conceded this point, and the article appeared with half a dozen instances of "spectra" and none of "spectrums".
For three other examples of usage issues for pluralization of borrowed words, see "Pseudo-Latin plurals", 4/23/2004, "Octopussies", 4/24/2004, and "What's the plural of syllabus?", 10/4/2010.
And there's probably a list Out There of contested (or contestable) borrowed-word plurals, but a quick search doesn't turn it up — any suggestions?
Jim Breen said,
May 24, 2022 @ 7:02 am
My shudder-inducing word these days is "alum" (and its plural "alums"). To me, an alum is a form of alumin[i]um sulfate, and the word shouldn't be used to sweep away the issues with alumnus/alumna/alumnae/etc. As the good people at Nottingham wrote: "Avoid using these words as much as possible; use "graduate" or "postgraduate" instead (or "former pupil" if talking about a school)"
https://www.nottingham.ac.uk/brand/styleguide/?id=0619301e-c097-4456-af3c-f3c42c7803cd
KeithB said,
May 24, 2022 @ 8:16 am
Off topic, but pharyngula has an article about French poems that when read aloud, sound like English nursery rhymes:
https://freethoughtblogs.com/pharyngula/2022/05/24/know-wonder-we-struggle-with-french-pronunciation/
J.W. Brewer said,
May 24, 2022 @ 8:17 am
@Jim Breen: Although dictionaries do not always track this nuance, "graduate" is not actually a perfect synonym for "alumn-X" in many contexts. Many universities (at least in the U.S.) are more than happy to recognize former students who never actually finished up and received a degree as "alumni" entitled to fully participate in "alumni" activities, attend reunions, and (most importantly!) be solicited for donations. I believe the formal rule at my alma mater is that anyone who completed three (out of the usual total eight) semesters of undergraduate study qualifies as an alumn[us/a] if he/she wishes to take advantage of that status.
The University of Texas has an "Ex-Students Association" (informally known as the "Texas Exes") rather than an "Alumni Association," and you can find similar usages for other institutions in Texas, but that seems to remain a regionalism.
mg said,
May 24, 2022 @ 8:40 am
@Jim Breen – I much prefer "alum" to the frequently seen mix-up between singulars and plurals. I hate seeing someone saying they are "an alumni" of someplace. As someone who is an alumna of several institutions, I'm perfectly happy to see the gender-neutral "alum".
The one that grates on me is millenniums instead of millennia.
Anthony said,
May 24, 2022 @ 8:56 am
I hope we never get to "phenomenons."
Leslie Katz said,
May 24, 2022 @ 8:58 am
Referendum: referenda? referendums?
Robert Coren said,
May 24, 2022 @ 9:07 am
@Anthony: Someone on a forum that I frequent used to write phenomenae as a plural for what he presumed was singular phenomena. I believe I was successful in setting him straight.
Criteria, on the other hand, seems to be getting increasingly entrenched as a singular.
Gregory Kusnick said,
May 24, 2022 @ 9:19 am
"Alum" seems no worse to me than "grad" for "graduate" or "memo" for "memorandum".
I did once have a boss who consistently used "memoranda" as singular.
Philip Taylor said,
May 24, 2022 @ 9:29 am
KeithB — I once presented our then Head of French with my own version of "Un petit d’un petit" —
Of course he said that it was nonsense French, but when the "joke" was explained to him he was not amused …
Don said,
May 24, 2022 @ 9:45 am
I was recently watching a series of tutorial videos about 3D modeling tool in which the instructor kept saying "vertice" (three syllables) as the singular of "vertices." He knew that it was wrong and would sometimes correct it to "vertex," but I guess "vertice" had been ingrained by habit.
Coby L said,
May 24, 2022 @ 9:48 am
One peculiarity of alumni/alumnae is that the Anglo-Latin pronunciation of the final syllable in the two words is exactly the opposite of the classical one.
Dennis Paul Himes said,
May 24, 2022 @ 9:52 am
The one which bothers me the most is "virii" as a plural of "virus". That's wrong on three different levels. Firstly, "virus" (meaning "slime" or "poison") is a mass noun without a plural. Secondly, it's neuter, so if it had a plural it would be "vira". Thirdly, if you misinterpret it as a second declension masculine (a reasonable mistake to make, actually) its plural would be "viri". There's no justification at all for the extra I, no matter what declension/gender you think it is.
J.W. Brewer said,
May 24, 2022 @ 9:54 am
It seems relevant that per the google n-gram viewer "data" has always been a more common word than "datum," way back to 1800 (and the same is true of e.g. "agenda" v. "agendum"), whereas by contrast "spectrum" has always been more common (albeit not nearly by the same order of magnitude as with the prior examples going the other way) than "spectra."
So my proposed hypothesis is that if Anglophones (esp those who have never formally taken Latin) are acclimated to the -um singular as the most-commonly-encountered form of the lexeme, they find the -a plural pretty easy to grasp, but if their primary contact is instead with the -a form (especially when the way it's used in English does not always make a "plural count noun" analysis intuitively obvious), they aren't nearly so good at back-forming the -um version.
Rodger C said,
May 24, 2022 @ 10:33 am
"Alum" seems no worse to me than "grad" for "graduate" or "memo" for "memorandum".
I take it you never deal with aluminum sulfate.
Rodger C said,
May 24, 2022 @ 10:35 am
The one that grates on me is millenniums instead of millennia.
What really grates in me is copyeditors changing "millennium" to "millenium."
rpsms said,
May 24, 2022 @ 10:38 am
Newton arguably "revived" the word spectrum (at least in scientific work) in "Optiks" and I note that he uses "spectrums." "Spectra" does not seem to appear at all in the printed work.
rpsms said,
May 24, 2022 @ 10:41 am
By "revived" I mean that Newton's use of the word for the visible spectrum seems to be the basis for the modern english speaker's metaphor of what a spectrum is.
Daniel Barkalow said,
May 24, 2022 @ 10:59 am
For me, it's definitely "spectra" if you're referring to the collection of frequencies a wave-like phenomenon can have contributions at (e.g., "the audio and radio spectra"), but I'm less certain when using "spectrum" to refer to the spectral density of a particular signal. It feels to me like this is one of the cases where extending a meaning can lose irregular plurals, and it's a question of which plural feels more natural if you're choosing one for "spectrum" anew. If you're talking about the defunct Illinois TV station "Spectrum" and the subsidiary of Charter Communications, I think you have to say they're completely different "Spectrums", and it would be very strange to use the irregular plural of the word in both of their names. For the analysis results, I could go either way.
M. said,
May 24, 2022 @ 11:26 am
Let us e n c o u r a g e regular plurals (spectrums, criterions, phenomenons, memorandums, alumnuses, appendixes, octopuses, millenniums, datums, viruses, vertexes, and so on) and that will put an end to the confusion, doubts, and conflicts. The fewer the number of irregularities, the better.
Andy Stow said,
May 24, 2022 @ 12:11 pm
In my job I deal with gas flows, and sometimes pipes go into or out of a larger chamber with low velocity, commonly called a plenum.
I like using "plena" rather than "plenums" as the plural, possibly just because nobody else does. Google ngrams shows some pretty wild ratios between the two.
David Marjanović said,
May 24, 2022 @ 12:11 pm
Lots of native English-speakers believe -ii is the plural of -us, or indeed the plural of anything Latin (leading to penii).
J.W. Brewer said,
May 24, 2022 @ 12:43 pm
Daniel Barkalow's point leads me to recall that if you grew up where and when I did, arthrous THE Spectrum (in a non-physics/engineering context) could only mean https://en.wikipedia.org/wiki/Spectrum_(arena). Because the referent was unique, it did not really need a plural form.
Check out FWIW this amazing folk-etymology: 'Lou Scheinfeld, former President of the Spectrum, explained that the name "Spectrum" was selected to evoke the broad range of events to be held there. "The 'SP' for 'sports' and 'South Philadelphia,' 'E' for 'entertainment,' 'C' for 'circuses,' 'T' for 'theatricals,' 'R' for 'recreation,' and 'UM' as 'um, what a nice building!"'
SlideSF said,
May 24, 2022 @ 1:19 pm
@Phillip Taylor: I am reminded of a book my father used to love. Mots d'heures: Gousses, Rames
https://en.wikipedia.org/wiki/Mots_d%27Heures
Philip Taylor said,
May 24, 2022 @ 1:22 pm
"Lots of native English-speakers believe -ii is the plural of -us, or indeed the plural of anything Latin (leading to penii)" — Is that really true, David ? I could easily believe that "Lots of native English-speakers believe [that] -ii is the plural of -ius" ("genii", for example) but would be surprised if they really generalise it to just "-us". For me (a native English speaker), the default plural for latin-based nouns ending in "-us" is "-i" (e.g., "focus" -> "foci").
[(myl) See "Pseudo-Latin plurals", 4/23/2004, referenced in the original post.]
Alexander said,
May 24, 2022 @ 1:46 pm
@Roger C: Since one is /ˈæl.əm/ and the other is /əˈlʌm/, I have no problem.
David Morris said,
May 24, 2022 @ 3:38 pm
There is one Peanuts cartoon in which Linus tells someone he is making *igli* out of eggshells for a class project. In one blog post I noted that the plural of iglu (igloo) is igluit (but I didn't note the source, so I can't be sure about that; it makes sense, compare Inuit) and the plural of gangurru (kangaroo) is gangurru-ngay (North Queensland, Guugu Yimithirr language, as researched by Haviland 1979).
Jerry Packard said,
May 24, 2022 @ 3:42 pm
I like to play different tenni, especially table, but not lawn.
Gregory Kusnick said,
May 24, 2022 @ 4:01 pm
Pronunciation aside, I suspect the number of (non-contrived) sentences in which "alum" is genuinely ambiguous between chemistry and institutional affiliation is vanishingly small.
Rodger C said,
May 24, 2022 @ 4:17 pm
I suppose my perspective is that of someone who learned /ˈæl.əm/ many years before /əˈlʌm/, and the latter always trips me up.
Steve Morrison said,
May 24, 2022 @ 8:22 pm
@Don: I once had a math teacher who used “parenthese” to mean “parenthesis.”
As to “alum” meaning “former student, irrespective of gender”: why not spell it “alumn?”
slz said,
May 24, 2022 @ 9:39 pm
Astronomy strikes again! We use spectra. The plural comes up extremely frequently because when you do spectroscopy, you'll likely end with multiple spectra of the same object if you observe it multiple times. I have not read any such prescription in journal style guides however, so I assume this is something that everyone agrees on without explicit instruction.
In other tricky astronomy-related plural forms, the plural of "ephemeris" (a table listing predicted positions of an object in the sky over time) is "ephemerides".
Dara Connolly said,
May 24, 2022 @ 10:16 pm
University of Nottingham said:
================================
"Avoid using these words as much as possible; use "graduate" or "postgraduate" instead (or "former pupil" if talking about a school)"
================================
I wonder if this instruction would seem odd or confusing to American readers, for whom the word "school" may refer to a university.
Dara Connolly said,
May 24, 2022 @ 10:22 pm
And there's probably a list Out There of contested (or contestable) borrowed-word plurals
Is there a list of words for which the plural varies depending on context and meaning? Antenna and index are two examples that spring to mind. I had a colleague (in a commodity trading function) for whom the regular singular form of "indices" was "indice".
Chris Partridge said,
May 25, 2022 @ 2:17 am
When I worked on The Times in London the style guide ruled that if the word was used in Roman times the Latin plural should be used, otherwise English plural form.
It is said that a professor of mathematics at a British University sent out a memo saying he wanted to improve the maths skills in the Faculty with a series of "Sunday fora, symposia or colloquia."
Another lecturer responded "So the Professor expects us to spend our Sundays sitting on our ba doing sa."
Andreas Johansson said,
May 25, 2022 @ 2:37 am
Like Steve Morrison, I've long wondered why the gender neutral form isn't "alumn".
I've run into a number of native anglophones who appeared to believe -ii is the default Latin plural ending. I wonder where they get the idea – I can't think of many Latin plurals ending in -ii that are at all likely to occur in English text (genii would be one), while many more end in single -i.
Is it just that the -ii looks so exotic that it sticks in extra well in the mind?
Philip Taylor said,
May 25, 2022 @ 3:32 am
"if the word was used in Roman times the Latin plural should be used" — clearly I was not awake when I first read this, since I interpreted it as "if the word was set in Time Roman the Latin plural should be used", and the logic thereof completely escaped me. It was only on returning to the thread after my first coffee of the day that I finally understood Chris's real meaning.
Bob Ladd said,
May 25, 2022 @ 6:36 am
A further Brit/Am difference with regard to "graduate" [noun] is that it only refers to a *university* graduate. The AmEng phrase "high school graduate" is completely weird in BrEng (and conversely with the equivalent BrEng phrase "school leaver"). Also, BrEng tends to use "graduate" where AmEng would use "college graduate", and there are phrases like "graduate jobs" meaning "jobs appropriate for someone who has been to university", which may be obscure to AmEng speakers.
bks said,
May 25, 2022 @ 6:51 am
Jerry Packard, are you sure the plural of tennis isn't elevenis?
David Morris said,
May 25, 2022 @ 7:00 am
Victor Borge's 'Inflationary language': https://www.youtube.com/watch?v=WmpLUezDzoo (or search for that title)
Philip Anderson said,
May 25, 2022 @ 7:41 am
@J.W. Brewer
A further hindrance to identifying the singular of an -a plural is that this could be a Latin -um or a Greek -on. For instance criterion/criteria, and I have just found out that there is a word criterium (for which I hope the plural is criteriums).
Radii is also found – I guess that an -ii termination is seen as definitely Latin, whereas -i occurs in other languages.
Mark P said,
May 25, 2022 @ 8:12 am
Someone I worked with on a government program referred to his list of items to be discussed at a meeting as an agendum. We never saw or heard that usage any other place.
Robert Coren said,
May 25, 2022 @ 9:54 am
@Andreas Johansson: "I can't think of many Latin plurals ending in -ii that are at all likely to occur in English text (genii would be one)"
I suspect that radii is the principal culprit.
Jerry Packard said,
May 25, 2022 @ 11:33 am
bks@
Jerry Packard, are you sure the plural of tennis isn't elevenis?
Very good – I am not sure at all. Maybe 'elevenae/eleveni' if it inflects at both ends.
Philip Taylor said,
May 25, 2022 @ 2:44 pm
At my tennis club, all games stop for elevenses.
Bloix said,
May 25, 2022 @ 7:09 pm
My father, who began his career as a spectroscopist, said spectra.
David Dickey said,
May 25, 2022 @ 8:31 pm
I well remember finding Garner's English Usage recommending the cringe-inducing genuses (900 genuses of dinosaurs, say) over genera. I'm a biologist who has never seen or heard that usage in the wild. Some of these are hard though, and seem mostly to serve as a way to keep out the hoi-polloi. How many of you could guess the proper form for multiple snake/lizard dicks (singular hemipenis — there is a righty, and a lefty, for disputed reasons); it's hemipene.
Mark P said,
May 25, 2022 @ 11:50 pm
More and more often I hear the plural of process as process-eze, with the accent on the “eze”, as opposed to process-es, where the final “es” unaccented. It seems to be common among scientists.
maidhc said,
May 26, 2022 @ 2:27 am
Mark P: Surely one item to be discussed at a meeting is an agendum, and a list of such is agenda.
You can wow everyone at your next meeting by saying "Our first agendum is …"
Terry Hunt said,
May 26, 2022 @ 3:22 am
@ KeithB and SlideSF – I have the (hilarious) German version of this, Mörder Guss Reims: selected poems of Gustav Leberwurst ("transcribed and annotated" by John Hulme). I've been hunting for a copy of the French version in second-hand bookshops for years. (I don't buy online – that would be too easy: the thrill is in the chase.)
Peter Erwin said,
May 26, 2022 @ 6:35 am
@Dara Connolly
I had a colleague (in a commodity trading function) for whom the regular singular form of "indices" was "indice".
I have seen a tutorial discussion of the geometry of objects in a 3D graphics program (possibly Blender) that used "vertice" as the singular form of "vertices".
Peter Erwin said,
May 26, 2022 @ 6:40 am
And there's probably a list Out There of contested (or contestable) borrowed-word plurals, but a quick search doesn't turn it up — any suggestions?
I have occasionally come across "shamen" as the supposed plural for "shaman" — the (generally excellent) fantasy writer Martha Wells is fond of this. (Either that or she has been afflicted by some weirdly insistent and misguided proofreader.)
mg said,
May 26, 2022 @ 11:20 am
As a medical researcher, it drives me nuts seeing bacteria used as a singular. One bacterium, multiple bacteria!
Philip Taylor said,
May 26, 2022 @ 2:35 pm
"Shamen" — Until you mentioned it, Peter, I would have accepted this as a correct plural without hesitation. Now that I know that it is not, and that the most etymologically correct plural would be "shamanen", I still prefer "shamen" !
Rodger C said,
May 26, 2022 @ 2:46 pm
I've also seen "talismen."
Philip Taylor said,
May 26, 2022 @ 3:18 pm
Again, "talismen" would not have seemed wrong to me, had you not pointed it out as such. And yet "humen" seems (to me) clearly wrong. So what is it, I wonder, that allows some words ending in "-man" to take a false plural in "-men" without causing any internal warning bells, while other words ending in "-man" automatically cause internal warning bells if they are pluralised as "-men" ?
Philip Taylor said,
May 26, 2022 @ 3:47 pm
Just heard the presenter of today's PM programme on BBC Radio 4 use "referendums" twice — it really grated on my ears.
Craig said,
May 26, 2022 @ 3:50 pm
@Steve Morrison and @Andreas Johansson, for what it's worth, Merriam-Webster Online https://www.merriam-webster.com/dictionary/alum claims that the first appearance of "alum" as a clipping to refer to a former student of either gender was 1877.
I imagine that the n-less spelling simply reflects the choice to clip at the syllable break.
ktschwarz said,
May 26, 2022 @ 3:51 pm
David Dickey: multiple snake/lizard dicks (singular hemipenis — there is a righty, and a lefty, for disputed reasons); it's hemipene.
That was a typo for "hemipenes", right? Though the singular "hemipene" is also out there, analogous to "vertice" and "indice".
Robert Coren said,
May 28, 2022 @ 8:52 am
Along with shamen and talismen (neither of which I can remember seeing in the wild), there's ombudsmen. While the -man element is the equivalent of English man, this wouldn't be a proper Scandinavian plural, although the Wikipedia entry says that "the modern plural form ombudsmen of the English borrowed word ombudsman is etymologically reasonable".
Andreas Johansson said,
May 30, 2022 @ 2:46 am
@Robert Coren:
The Swedish plural is ombudsmän, but ombudsmen if it existed would be pronounced identically.
"Shamen" I've seen often enough, mostly in the context of fantasy books and games – perhaps never in the context of real-world practitioners. Don't think I've seen "talismen" before.
A funny case is "Normans" v. "Northmen" – etymologically identical, but with differently formed plurals.