Ligatures, Diphthongs and Supernovae

At the weekend I noticed a nice article by John Butterworth on his Grauniad blog about where Gold comes from. Regular readers of this blog (Sid and Doris Bonkers) know that I am not at all pedantic but my attention was drawn to the plural of supernova in the preamble:


I have to confess that I much prefer the latin plural “supernovae” to the modernised “supernovas”, although most dictionaries (including the One True Chambers) give these both as valid forms.  In the interest of full disclosure I will point out that I did five years of Latin at school, and very much enjoyed it…

When I tweeted about my dislike for supernovas and preference for supernovae some replied that English words should have English plurals so that supernovas was preferred (although I wonder if that logic extends to, e.g. datums and phenomenons). Others said that supernovae was fine among experts but for science communication purposes it was better to say “supernovas” as this more obviously means “more than one supernova”. That’s a reasonable argument, but I have to admit I find it a little condescending to assume that an audience can cope with the idea of a massive star exploding as a consequence of gravitational collapse but be utterly bewildered by a straightforward latin plural.

One of the reasons I prefer the Latin plural – along with some other forms that may appear archaic, e.g. Nebulae – is that Astronomy is unique among sciences for having such a long history. Many astronomical terms derive from very ancient sources and in my view we should celebrate this fact because it’s part of the subject’s fascination. That’s just my opinion, of course. You are welcome to disagree with that too.

Anyway, you might be interested to know a couple of things. One is that the first use of “super-nova” recorded in the Oxford English Dictionary was in 1932 in a paper by Swedish astronomer Knut Lundmark. This word is however formed from “nova” (which means “new” in Latin) and the first use of this term in an astronomical setting was in a book by Tycho Brahe, published in 1573:

Brahe_book(I’ll leave it as an exercise to the student to translate the full title.)

Nowadays a nova is taken to be a much lower budget feature than a supernova but the “nova” described in Tycho’s book was was actually a supernova, SN1572 which he, along with many others, had observed the previous year. Historical novae were very often supernovae, in fact, because they are much brighter than mere novae. The real difference between these two classes of object wasn’t understood until the 20th Century, however, which is why the term supernova was coined much later than nova.

Anyway, back to pedantry.

A subsequent tweet from Roberto Trotta asserted  that in fact supernovae and supernovas are both wrong; the correct plural should be supernovæ, in which the two letters of the digraph “ae” are replaced with a single glyph known as a ligature. Often, as in this case, a ligature stands for a diphthong, a sort of composite vowel sound made by running two vowels together.   It’s one of the peculiarities of English that there are only five vowels, but these can represent quite different sounds depending on the context (and on the regional accent). This  means that English has many hidden diphthongs. For example,  the “o” in “no” is a diphthong in English. In languages such as Italian, in which the vowels are very pure, “no” is pronounced quite differently from English. The best test of whether a vowel is pure or not is whether your mouth changes shape as you pronounce it: your mouth moves as you say an English “no”, closing the vowel that stays open in the Italian “no”…

So, not all diphthongs are represented by ligatures. It’s also the case that not all ligatures represent diphthongs. Indeed some are composed entirely of consonants. My current employer’s logo features a ligature formed from the letters U and S:


The use of the ligature æ arose in Mediaeval Latin (or should I say Mediæval?). In fact if you look at the frontispiece of the Brahe book shown above you will see a number of examples of it in its upper-case form Æ. I’m by no means an expert in such things but my guess is that the use of such ligatures in printed works was favoured simply to speed up the typesetting process – which was very primitive – by allowing the compositor to use a single piece of type to set two characters. However, it does appear in handwritten documents e.g. in Old English, long before printing was invented so easier typesetting doesn’t explain it all.

Use of the specific ligature in question caught on particularly well in Scandinavia where it eventually became promoted to a letter in its own right (“aesc”) and is listed as a separate vowel in the modern Danish and Norwegian alphabets.  Early word-processing and computer typesetting software generally couldn’t render ligatures because they were just too complicated, so their use fell out of favour in the Eighties, though there are significant exceptions to this rule. Latex, for example, always allowed ligatures to be created quite easily. Software – even Microsoft Word – is much more sophisticated than it used to be, so it’s now not so much of a problem to use ligatures in digital text. Maybe they will make a comeback!

Anyway, the use of æ was optional even in Mediaeval Latin so I don’t think it can be argued that supernovæ is really more correct than supernovae, though to go back to a point I made earlier, I do admit that a rambling discussion of ligatures and diphthongs would not add much to a public lecture on exploding stars.


17 Responses to “Ligatures, Diphthongs and Supernovae”

  1. Bryn Jones Says:

    Yes, I agree: the word should be supernovae, not supernovas or supernovæ.

    I too studied Latin at school for five years, at a comprehensive school, of course. Admittedly, I’ve forgotten it all now.

  2. Anton Garrett Says:

    You are certainly not being oeconomical with the truth.

  3. ä, ö, and ü also originally were a type of ligature, with an e written over the corresponding letter; this degenerated to two dots. Similarly, the letter å originated as an a with an o over it. (This is the usual way to write the o sound in Scandinavian languages, o sounding more like u, u sounding like ü and y something between ü and i.)

    • telescoper Says:

      Interestingly the symbol “&” was also originally a ligature, formed from the letters “e” and “t”…

    • I am mildly bemused by the modern trend in German to accept only umlauted forms and reject thje “oe”, “ae” and “ue” spellings — which I find are actually quite common in older texts. (That said one German aquantance of mine is pedantic about using only US-ASCII).

      But my bemusement is only because for dictionary writers and other quasi-officials of the English language, the only legitimate trend is towards less strictness, Hence “supernovas”. Why insist on one correct spelling when you can two or more it was good enough for William Shaksper

      • What do you mean by the “trend in German to accept only umlauted forms”? Either spelling is “anything goes” (of course, in practice, too strange spellings will not find favour) or there is some official orthography. If there is an official orthography, it should be followed. Thorn and eth aren’t used in English anymore.

        The umlauts have almost become fully fledged letters. As a result, the name “Bueb” is pronounced with two syllables, since otherwise it would be written with an Umlaut. In other words, the diaeresis over the “e” is unnecessary. (Confusingly, this looks almost exactly like an umlaut sign, though the latter is often set closer to the letter).

        Sure, they exist in older texts, but I don’t know who “rejects” them.

        Even though spelling in German has changed about the same amount since Shakespeare’s day, it is customary in German to publish old literature in the modern spelling, which isn’t the case in English.

        US-ASCII is the lowest common denominator and is still advisable if you have a 7-bit email connection or whatever.

  4. In Swedish, Swedish words ending in a form the plural in or. Foreign words are often treated similarly, so it is supernovor.

  5. “A subsequent tweet from Roberta Trotta”

    Roberto, surely.

    Last Sunday, I was at a concert which featured Robert Crowe (along with a lute player and a harpsichord player). If one used only one’s ears, one could be forgiven for guessing the incorrect gender.

  6. I did a lot of thinking about plural forms when I was writing a text. Ultimately I concluded that while it is fun to write ‘supernovae’ and ‘radii’ and other Latinate (or other-language) plurals, these are nevertheless words that have been firmly taken into English. And there’s not much reason to make English harder than it has to be, especially in technical or scientific writing. There’s already enough stuff to think about, and there’s a probably higher proportion of people for whom English isn’t a first language reading the technical or scientific prose. So it’s probably kinder to the reader to make as few exceptions to normal pluralization (or, where appropriate, verb conjugation) rules.

    Also, pluralizing ‘virus’ just gets people angry with you.

    • telescoper Says:

      It would make me angry if anyone used “viri” as the plural of virus as that is incorrect in Latin. The latin word virus has no plural, as it is a singular mass noun in its original meaning (ooze or slime). The only sensible plural in English is viruses.

      Viri in Latin is actually the plural of “vir”, a totally different word..

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: